Privacy Structure Analysis And Path Planning Of Application Cooperative Network Based On Heuristic Algorithm

In recent years,with the explosion of Internet and network data,the studies on collaborative networks have increasingly become a research hotspot,which have been attracted focus from researchers in all over the world.The collaborative network can be used to represent the real-world systems.For instance,the urban traffic system can be represented as a multiplex network with two layers,in which one layer represents ground bus system,and the other one represents the underground subway system.The real-world interpersonal network can be modeled as a singed network,in which positive(negative)edges represent the friendly(hostile)relationships between individuals.The geographic survey system with geographic location information can be represented as a spatial information network.In this paper,we mainly study the applications on the signed collaborative networks and the spatial collaborative networks.Signed networks are used to represent complex interpersonal relationships with potential conflicts and tensions.The potential conflicts in signed networks can be easily revealed by computing their global structure balance(GSB).However,the existing studies for GSB neglect the privacy of link information,and thus the personal privacy information is easy to be exposed.Here,we study the GSB problem in signed networks,considering the privacy preservation of personal information.Spatial information networks are used to represent the spatial systems with specific spatial geographic location information.Unmanned Aerial Vehicle(UAV)is often used to collect geographic information.However,the limited energy and endurance of UAV result in that UAVs are hard to be applied for some applications with wide distribution of geographic information.Here,we propose an unmanned air/ground cooperative system to collect the geographic information,which comprehensively utilizing the advantages of UAVs in fast speed,and UGVs in high endurance and loading ability.Moreover,we propose two heuristic algorithms to solve the GSB problem and the path planning problem,considering the tradeoff the computational complexity and effectiveness of the algorithms.Specially,we mainly study the following two aspects.By analyzing the existing privacy protection strategies,we propose a signed network privacy protection framework based on random perturbation policy and homomorphic encryption technology.This framework is composed of the sides of clients and servers.Clients can first use their own mobile devices and some encryption strategies provided by the servers to encrypt their own privacy information,and then send the encrypted data to the servers for data analyses.Homomorphic encryption technology is adopted on the server side,which enables statistical analysis of user data and decryption of the final results without obtaining the original data of each client.Moreover,we design a fast greedy algorithm to calculate the global structure balance of the encrypted signed network.Specifically,this algorithm aims to find a best clustering division with a minimum number of unbalanced edges.Experiments results on synthetic signed networks and real-world signed networks show the effectiveness of the proposed framework and algorithm in the privacy protection and the global structural balance computation.By considering the application of the detection of illegal building information in cities,we propose a novel unmanned air/ground cooperative system to collect geographic information with an optimal path planning.Moreover,we propose a two-level memetic algorithm to find the optimal path planning for the unmanned air/ground cooperative system.Experiments on both simulated geographic data and real-world geographic data show the advantage of the proposed algorithm over the classical algorithm on the path planning for the system to complete the detection of non-standard buildings in cities.